Defence CommitteeWritten evidence from Steven Haines, Professor of Public International Law, University of Greenwich MA, PhD, LLM. Professor of Public International Law, University of Greenwich (since 2012). Formerly: Geneva Centre for Security Policy (2008-2012) as Head, Security and Law Programme; Royal Holloway College, University of London (2003-2008) Head of the Department of Politics and International Relations and Professor of Strategy and the Law of Military Operations; Royal Navy (1971-2003) with operational service on UN maritime embargo operations (1973), with the Northern Ireland Security Forces (1977-82), NATO(KFOR) in Pristina, Kosovo and the UK JTFHQ in Freetown, Sierra Leone (2001). Service in Ministry of Defence (1995-2003): on Naval and Central Policy Staffs; Executive Board and Head of Project and Research Management, Joint Doctrine and Concepts Centre, Shrivenham. Author: British Defence Doctrine (2001); RN Handbook on the Law of Maritime Operations (2003). Chairman of the Editorial Board, UK’s official Manual of the Law of Armed Conflict (2004). Elected Member, International Institute for Humanitarian Law, San Remo, Italy. Visiting appointments: Cranfield University Security Studies Institute (2000-2010); St Antony’s College, Oxford (2001); Geneva Institute for International Humanitarian Law and Human Rights (2010-2012); International Security Programme, Federal Institute for Technology, Zurich, Switzerland (2010 to present); University of Oxford Changing Character of War Programme (2012-2013).

I offer comments on the following:

The Legality of Military Intervention.

The Influence of Human Rights Law on the Application of Force.

Specific Weapons Law Issues (Autonomous Weapons and Drones)

The Legality of Military Intervention

In recent years, and especially following the controversial decision to invade Iraq in 2003, there has been a significant amount of debate about the constitutional process by which HMG decides to deploy military force overseas for operational purposes in which the use of force is likely to be required. Following the Iraq experience, the current Prime Minister (PM) made it clear that he required an appropriate UN Security Council mandate for the intervention in Libya and Parliament was engaged prior to the deployment of the Armed Forces. In the case of Syria, no UN mandate was likely and the PM chose to go to Parliament for a decision that eventually led to the UK not mounting any military operation. Interestingly, President Obama also referred deployment to Congress leading to a US decision not to intervene. In fact and in law, neither the PM nor the US President were under any legal or constitutional obligation to hand such decisions to their respective legislature (the US War Powers process would not have necessitated reference to Congress for the type and duration of deployment then envisaged)

The core question for the UK is about where responsibility for decision-making on military deployments should rest—with the Executive under the Royal Prerogative or with Parliament. Prior to the recent Libya and Syria experiences, the previous Labour Government, in the wake of Iraq, had seriously considered introducing “War Powers” measures. It was clearly motivated by a desire to avoid the degree of legal controversy that followed the decision to invade Iraq. In October 2007, the Ministry of Justice published a consultative document with proposals for the introduction of War Powers mechanisms to replace the existing Royal Prerogative.2 I submitted a response to that CP in January 2008 and was subsequently called to give evidence before the Joint (Parliamentary) Committee on the Draft Constitutional Renewal Bill on 13 May 2008.3 For information, I attach to this submission a copy of my January 2008 response to the Ministry of Justice CP.4

In brief, in 2008 my position on “war powers” was that the decision to deploy the Armed Forces should remain a function of the Executive under the Royal Prerogative. The involvement of Parliament in decision-making is potentially extremely problematic in operational terms. Issues that are difficult to resolve include the setting of a threshold of scale or category of deployment for parliamentary engagement, and the extent to which Parliament remains involved in decision-making during operations. I will resist rehearsing the various issues that were addressed during the discussions in 2008 (a number of which I address in my attached response to CP26/07) but, in the end, various concerns proved sufficiently problematic for HMG to decide that no War Powers provisions should go forward into legislation—at least, not at that point.

The constitutional situation today arguably remains as it was in 2008 and previously, with HMG, under the Royal Prerogative, being in the legal position to take decisions on Armed Forces deployment without reference to Parliament. I say “arguably” because it is also arguably the case that the decision by the PM to refer the situation in Syria to Parliament is some measure of precedent. Under our law, constitutional change is frequently achieved not through legislation but through shifts in practice. It is an interesting question whether or not the current PM’s decision to refer Syria to Parliament has set a precedent that subsequent PMs will find it difficult not to repeat. One suspects this is the case but, as with all such constitutional shifts, we must wait for subsequent experience to either confirm a shift in that direction or mark the Syria decision out as an exceptional departure from a constitutional norm.

Whatever the future holds in that respect, I remain convinced that giving Parliament a role in decision-making of this sort would be unfortunate. It is the role of HMG in our constitution to decide within the law, with Parliament determining what the law should be. Clearly this means that if Parliament decides through legislation that military deployment decisions must be made by Parliament, then future governments will be obliged to comply with that. However, I believe this would grant Parliament inappropriate executive responsibility for matters that it is not appropriately configured to shoulder.

That is not to say that Parliament should not be engaged. As I pointed out in my response to CP26/07, in fact successive governments over many years have engaged with Parliament over major deployments of the Armed Forces. I well remember the significance of the Falklands debate in the House of Commons in 1982. It would have been inconceivable for HMG to have gone to war in the South Atlantic without broad political support expressed in the two chambers in Parliament. This will always be so. Indeed, even the invasion of Iraq was subject to parliamentary debate. It is easy to forget that there was broad (though certainly not unanimous) political support for that deployment. The proposed intervention in Syria was not a deployment in anything like that league, however. In military terms, if it had gone ahead it would have amounted to a relatively short operation with little or no deployment by Armed Forces personnel into Syrian territory. One imagines that submarine launched cruise missiles would have been a significant feature of the assault on Syria and that the attack would have lasted a matter of hours rather than days. There will always be those who warn against so-called “mission creep” but this was a very low risk indeed in relation to Syria. That is not to say that the attack should have occurred.

What we must try to achieve in our decision-making for military deployment is a process that leads to compliance with the law—and not just domestic law. The use of force within the international system is something that is covered by international law and it is vital that the UK acts within that law. In 2003, it was arguably the case that the invasion of Iraq was contrary to international law. As an international lawyer, I believe it was unlawful—and the bulk of international legal opinion agrees with me. I have to say that the legal case put forward by Lord Goldsmith at the time was logical, well thought through—rigorous, indeed—and for some it was persuasive. However, it would not have prevailed if the matter had ever been put through a judicial process during which international law was applied. The opposing argument was much stronger and far more convincing. Moving on to Syria, if the PM had not gone to Parliament at all, or had given it a vote and then ignored the outcome, a decision to strike Syria would have been contrary to international law. Such a strike required a clear UN Security Council mandate. The humanitarian intervention argument—used very powerfully at the time of NATO’s action against Serbia over Kosovo in 1999—is now a weaker one given the way in which the law has developed since 1999.5 It is one of the major dilemmas of the international security system that a veto by a Permanent Member of the Security Council can prevent a lawful military response to a humanitarian crisis, by simply rendering any military intervention unlawful. The PM decided not to go against the wishes of Parliament and, in so doing, complied with international law. That, however, was not his intention and he would have gone ahead and ordered military action if Parliament had not said that he should not. This raises an interesting question about the law, which I believe is the crucial issue for the Committee to reflect upon.

At the moment, there is no domestic UK law or process preventing a Prime Minster, with Cabinet support, from mounting military operations that would be contrary to international law. Neither Tony Blair over Iraq nor David Cameron over Syria—if he had ignored Parliament—could have been prosecuted in a British court for a breach of the law governing the use of force at international law. As an international lawyer, committed to British compliance with the law in that sense, I believe this to be profoundly regrettable. Instead of a process that involves Parliament becoming involved in decision-making about military deployments, I would prefer to see the Supreme Court charged with assessing the legality of resort to force under international law. If a PM knew that any decision he or she made about military deployment would be subject to rigorous subsequent legal scrutiny, that knowledge would act as a break on action likely to be adjudged unlawful. In the past I thought that Parliament ought to be involved in the post-deployment scrutiny process, but I have shifted and now believe this is a role that ought to be performed by the Supreme Court in order that a judgement is arrived at based on the law and not influenced—or suspected of being influenced—by other wider political considerations.

Aside from being an academic international lawyer, I am a retired officer in the British Armed Forces who, in an active career of over thirty-two years, deployed on operations in Northern Ireland, the Balkans and Sierra Leone. At no point in my Service career until the year of my retirement in 2003 did I ever have cause to question the legality of what the UK was using its Armed Forces for. The bulk of those who serve in uniform like to know that what they are doing and what their country is using them for is right and legitimate. In 2003, over Iraq, there were severe doubts in that regard. This is why Admiral Boyce, as Chief of the Defence Staff, requested clarification from the PM that the invasion would be legitimate—in doing so he was not seeking to clear himself of any legal responsibility for a bad decision but was instead anxious to put the minds of those men and women about to invade a foreign country that they would not be breaking any law. That is important—vitally important, for their motivation and their peace of mind.

My appeal to the Committee is to take advantage of this opportunity to think seriously about how the decision-making process for military deployment can be rendered fit for application in the wider world—not to place decision-making in the hands of Parliament but to create an effective process of legal scrutiny of those decisions made by the Executive under the Royal Prerogative.

The Impact of Human Rights Law on the Application of Force

There has been a great deal of comment in recent years about the ways in which Human Rights Law (HRL) is assumed to influence the manner in which military operations are conducted. Traditionally, in armed conflicts (what used to be described as “war”) during which the Geneva Conventions and the full range of the Laws of Armed Conflict (LOAC) apply, combatants would be constrained merely by the limitations imposed on targeting by the rule of distinction as between enemy combatants and military objectives, on the one hand, and civilians and civilian objects, on the other. Very simply, if an individual is identified as an enemy combatant (and he is neither wounded and hor de combat nor in the process of surrendering), he can be targeted, no matter what he is doing at the time. So, for example, even an apparently unarmed enemy combatant who seems to be running away from the fighting, remains a legitimate target; he can even be shot in the back while appearing to pose no immediate threat to anyone. His mere identity as an enemy combatant renders him a legitimate target, no matter what he is doing at the time he is being targeted. This contrast with situations short of armed conflict in which the targeting of individuals is driven not by their identity but by their actions. In law enforcement situations, the guiding principle is to employ the minimum force necessary to achieve the objective of enforcing the law. It is not the identity of a person that determines their vulnerability to attack but their actions at the time they are confronted. Even if they are employing force themselves, the legitimate response must take account of the twin requirements of proportionality and necessity. The employment of excessive force would be unlawful.

There is a common perception these days that British servicemen deployed on operations are, for reasons of human rights compliance, being obliged to apply force using what is in effect the human rights based law enforcement paradigm rather than that permitted under LOAC, or the armed conflict paradigm. To a significant degree, however, this perception is flawed. There was a time when Rules of Engagement (ROE) were promulgated for the application of force in situations short of armed conflict—in circumstances associated with law enforcement. If in armed conflict, opening fire rules were in accordance with the armed conflict paradigm and no ROE were necessary. This is decreasingly the case today for a reason associated with the complexity of the operational environment and the need to tailor the application of force to the precise circumstances encountered. In Afghanistan, for example, NATO has been dealing with a situation of insurgency demanding a sophisticated approach to the conduct of operations. For operational reasons rather than legal ones (related to the demands of HRL), ROE have been imposed to tailor the application of force in a manner consistent with operational objectives. Excessive use of force would, in some circumstances, be counter-productive. ROE are imposed, not to ensure compliance with the law (either LOAC or HRL), but to ensure that what happens at the tactical level is consistent with strategic objectives and directions. ROE are a command and control mechanism and not a means of ensuring compliance with the law. Importantly, they are not a means of imposing HRL standards on the battlefield although, if one wanted to do that, ROE would be the most appropriate mechanism for achieving that.

I do not believe that HRL is undermining the effectiveness of military operations by the imposition of excessive restraints on the application of lethal force. For one thing, it is simply not correct to maintain that HRL protects the life of opposing belligerents in a way that prevents the Armed Forces achieving their objectives. HRL does not contain an absolute right to life. Rather, it asserts the right of all not to be deprived of their life in an arbitrary fashion. In some circumstances, it may be absolutely necessary for a person committing or about to commit an offence to be killed; it may be the minimum appropriate force that it is necessary to use to prevent an offence being committed. If they are killed in those circumstances, there will have been no breach of HRL because, while they may have been killed, their death will not have been the result of an arbitrary decision to kill. Rather, it will be the result of a calculated decision to apply force in an appropriate way for the circumstances faced. In regions of Afghanistan where there is intense combat (Helmand, for example), I have not heard that HRL is having undue influence on opening fire regulations. If there are restrictions that require restraint greater than that required by basic LOAC, it is being imposed for other operational (and not legal) reasons. I have heard people claim that HRL is problematic; such claims are, however, almost certainly ill informed and may even be influenced by a particular predisposition to be suspicious of anything to do with human rights. I must, though, caveat what I have just said by stressing that I have not myself experienced the circumstances to which I allude. My recommendation to the Committee would be to interview experienced commanders and soldiers with recent experience of operations in Afghanistan to establish whether or not what I have just said is correct. For those interviews to be of value, however, Committee members will themselves need an adequate understanding of the purpose of ROE and the different demands of the armed conflict and law enforcement paradigms.

Specific Weapons Law Issues

There are two types of weapon on which I believe I can provide brief useful comment. I do so as a specialist in Weapons Law who, when serving in the Ministry of Defence, was one of the three tri-Service military staff officers responsible for conducting reviews of new weapons, means and methods of warfare for the UK. These reviews were conducted in accordance with Article 36 of the 1977 Additional Protocol I to the 1947 Geneva Conventions, which requires all states to ensure that all weapons for use in armed conflict are compliant with the weapons law element of LOAC. The comments I make below are principally related to legality under LOAC

Un-Manned Aerial Vehicles (UAVs) aka “Drones”

UAVs are, quite simply, aeroplanes. The fact that they are un-manned and remotely operated does not alter that basic fact in any way. Aeroplanes are not unlawful; UAVs are not, therefore, in and of themselves unlawful. UAVs operating exclusively for reconnaissance, data gathering and intelligence purposes are not weapons. It is only when a UAV is weaponised that it becomes a weapon and is required to be compliant with LOAC weapons law. Even then, as long as the weapon it is carrying is itself lawful (not subject to a ban under the Certain Conventional Weapons Convention, for example), the UAV will be compliant with the law. That is not to say that it cannot be operated in an unlawful manner or for unlawful purpose. All weapons can be put to unlawful purpose; UAVs are no different from other weapons in that respect.

Weaponised UAVs are frequently assumed to be operating in an unlawful manner or for unlawful purpose. The claims made against UAV operations currently fall into two forms in general.

First, it is frequently claimed that they are operating where they should not be operating and where their operators have no legal right to deploy them. UAVs operating in the border areas between Afghanistan and Pakistan are a particular focus for criticism of this sort. Their presence in Pakistani airspace and their targeting of individuals on the ground in Pakistan is a legal issue to do with the international law relating to the use of force and the legitimacy of intervention. If the UAV’s operating authority has no permission from Pakistan to deploy the aircraft into Pakistani airspace, their presence there will be, prima facie, unlawful. If, on the other hand, permission has been granted by the Pakistani authorities, the presence of UAVs in Pakistani airspace will not be unlawful.

Second, it is claimed that the targeting of individuals on the ground from UAVs is unlawful, either because it represents the arbitrary killing of individuals without trial or because attacks result in the deaths of innocent civilians who are “collateral” casualties in an otherwise legitimate attack (if, indeed, the attack itself is legitimate). The unintentional killing of civilians during lawful attacks on enemy combatants or those directly participating in hostilities, is not necessarily unlawful. LOAC recognises that some civilian casualties may occur but that if they do there may be no breach of LOAC of the deaths are proportionate to the direct military advantage gained from the attack on the lawful target. If UAVs are launching attacks under the armed conflict paradigm, a number of civilian casualties, while regrettable, may be permitted. A lot will depend on the precise circumstances of each attack that results in civilian casualties. If the attacks are carried out under the law enforcement paradigm, there is no allowance based on proportionality and all civilian deaths will be prima facie unlawful. Importantly, however, the deployment of UAVs and their use for targeting identified combatants is generally more likely to achieve the objective with minimum civilian casualties than any strike mounted in any other way. UAVs allow for a more thorough compliance with the principle of distinction and result in few civilian casualties than would be the case if alternative means were used to target enemy combatants.

Autonomous Weapons

There are already weapons systems that are capable of being triggered automatically on the detection of a “target” that meets the features to which the weapon is programmed to respond. Good examples are the “close-in weapons systems” deployed at sea to protect ships from surface to surface missile attack, the Dutch Goalkeeper and the US Phalanx systems being the best known. Both use radar to detect incoming missiles, a fire control system to acquire the target, and a multi-barrelled gatling-gun to engage. These are not new systems, both becoming operational in about 1980. In one sense they are autonomous, in that they are programmed and left to detect and engage incoming targets without the need for human intervention. Indeed, human intervention would be too slow to deal with the incoming threat, so these systems are a defensive necessity against sea-skimming missiles. The systems’ reaction time is in the order of five or six seconds. These are not truly autonomous weapons, however.

True autonomy would take a weapons system to an altogether higher level of decision-making. Indeed, it would be capable of taking the sorts of decisions that currently require human intervention. Rapid reaction and automatic engagement in response to predetermined target features is one thing. The ability to seek out potential targets, assess their status and then to decide when to engage, represent a combination of capabilities that we do not as yet possess. Nevertheless, there are researchers working with that objective in mind. One notable figure in the search for true autonomy is Dr Ron Arkin, who believes that the development of computer technology will soon take us to the point at which it will be possible to fit weapons systems with sufficiently sophisticated software for the system to make targeting and engagement “judgements” without the need for a “man in the loop”. Indeed, in his published work, Arkin goes so far as to suggest that, given time constraints and data complexity, computer technology will be capable of achieving better and more humane compliance with LOAC than is possible using human judgement.6 Computers making the decisions would be strictly objective, unaffected by emotional factors, and incapable of deliberately targeting anything other than those meeting the strict features programmed into the fire control system within the weapon.

In November 2012, Human Rights Watch, in cooperation with the Harvard Law School’s Human Rights Clinic, published a pamphlet that seems to have set the immediate future agenda for those involved in the development of the law governing autonomous weapons.7 The pamphlet was a critique of so-called “Killer Robots” and dealt with the likely future emergence of wholly autonomous weapons, with computer technology making decisions on attack. One can, with justification, regard the Human Rights Watch initiative as a direct response to the work of those, like Arkin, who believe that at some point the human can be taken out of the decision-making loop. The immediate contrary response has come principally from the International Law Department at the US Naval War College. Jeffrey Thurner, a member of the War College Faculty, published an article in the influential Joint Force Quarterly in December 2012. This was followed by a substantial legal response in the Harvard National Security Journal from Michael N Schmitt, the Chair of the College’s International Law Department.8

Autonomous weapons are in the sights of civil society activists; government experts are preparing their response. Arguments are not merely about the law but also moral and ethical issues to do with the nature of conflict. There is a reputable philosophical case to be made that decisions to kill should always be taken by a moral actor—a soldier or military commander. The Human Rights Watch initiative sets out to stop technological development before it reaches the point of producing true autonomy. It seeks to have future research on autonomy declared contrary to weapons law, which it certainly is not at present. Nevertheless, Article 36 of Additional Protocol I requires legal checks to be made on developing weapons technology at various stages in its development, so there is a legal requirement to consider legality even arguably at the present stage of development.

Key issues to address in the legal assessment of any potential future weapon will be its ability to achieve compliance with the principle of distinction, which demands that only combatants be targeted. If the technology is intended for direct use against combatants, will the technology be capable of distinguishing them from civilians? While this seems unlikely, given the contemporary conflict environment in which insurgents are on first sight indistinguishable from the mass of the people, it may be possible to programme a weapon to assess over time the behaviour of a potential target and then to engage once a pattern of behaviour has been established beyond a particular point. Is this morally and ethically acceptable? This is the key question in the debate of autonomy.

My own approach is to monitor the development of the relevant technology, to keep engaged with those working in that field and to input a legal opinion when one seems appropriate. I am cautious about computers making decisions that replace the moral judgements of those doing the fighting or commanding military forces. To me, war/armed conflict is a profoundly moral endeavour and any attempt to take human judgement out of the equation needs to be met with concern. But I am reluctant at this point to rule out something which may have the ability to make targeting more effective and more consistently compliant with LOAC than fallible human judgement can be over time. One responsible approach may be to delay reaching a conclusion about the issues at stake. As Schmitt has noted:

“No such weapons have even left the drawing board. To ban autonomous weapons systems altogether based on speculation as to their future form is to forfeit any potential uses of them that might minimize harm to civilians and civilian objects when compared to other systems in military arsenals.”

It is far too early to predict the outcome of this debate but a wide-ranging series of legal, moral and ethical arguments are in prospect.

January 2014